Akash Hadagali Persetti

Shipping production agentic AI systems on AWS | LangGraph, FastAPI, Bedrock | MS CS ’26, Indiana University | Open to New Grad AI Engineer & SWE roles

Joined on Jun 29, 2026 https://akashpersetti.hashnode.dev

Jul 20

My eval leaderboard published a negative retry count

#llm #python #aws #architecture

7 min read

Want to connect with Akash Hadagali Persetti?

Create an account to connect with Akash Hadagali Persetti. You can also sign in below to proceed if you already have an account.

Create Account

Already have an account? Sign in

Akash Hadagali Persetti

Jul 18

I Ran terraform validate Inside a Lambda. Two Things Broke.

#terraform #aws #docker #llm

6 min read

Akash Hadagali Persetti

Jul 15

The cost of a serverless agent isn't the Lambda. It's the loop.

#aws #serverless #langgraph #ai

5 min read

Akash Hadagali Persetti

Jul 12

Four rival LLMs, zero consensus: designing an MCP panel

#mcp #llm #python #aws

5 min read

Akash Hadagali Persetti

Jul 9

I grade my AI twin on real traffic, not just an offline test set

#aiops #aws #rag #ai

5 min read

Akash Hadagali Persetti

Jul 8

I built an eval harness for my own AI, and it caught my digital twin lying

#llm #ai #python #aws

6 min read

Akash Hadagali Persetti

Jul 3

Building a worker-evaluator retry loop in LangGraph (and where it bites you on Lambda)

#langgraph #terraform #lambda #agents

4 min read

Akash Hadagali Persetti

Jun 30

I stopped putting AWS keys in GitHub Secrets. Here's what I do instead.

#aws #terraform #githubactions #iam

5 min read

Akash Hadagali Persetti

Jun 29

Why my LangGraph agent throws away its checkpoint on every request

#agents #architecture #aws #serverless

4 min read

DEV Community

Akash Hadagali Persetti

Badges

Writing Debut

My eval leaderboard published a negative retry count

Want to connect with Akash Hadagali Persetti?

I Ran terraform validate Inside a Lambda. Two Things Broke.

The cost of a serverless agent isn't the Lambda. It's the loop.

Four rival LLMs, zero consensus: designing an MCP panel

I grade my AI twin on real traffic, not just an offline test set

I built an eval harness for my own AI, and it caught my digital twin lying

Building a worker-evaluator retry loop in LangGraph (and where it bites you on Lambda)

I stopped putting AWS keys in GitHub Secrets. Here's what I do instead.

Why my LangGraph agent throws away its checkpoint on every request